troggle-unchained/databaseReset.py

import json
import locale
import os
import resource
import sys
import time

import settings

""" Command-line utility for loading cave data files into troggle's database.

The command line options select which combination of classes of data will be imported, 
e.g. cave data, logbook data, cavers (people) data. The set of imports requested are put 
into a job queue in a valid order, toegether with any necessary initialisation if it is
a complete reset, and the queue is then executed.

In future all these functions may be moved to a control panel webpage running within the
troggle application.
"""
print(" - settings on loading databaseReset.py", flush=True)

os.environ["PYTHONPATH"] = str(settings.PYTHON_PATH)
os.environ.setdefault("DJANGO_SETTINGS_MODULE", "settings")

print(" - settings on loading databaseReset.py")

import django

print(f" - Memory footprint before loading Django: {resource.getrusage(resource.RUSAGE_SELF)[2] / 1024.0:.3f} MB")
try:
    django.setup()
except:
    print(
        " ! Cyclic reference failure. Can occur when the initial db is empty. Fixed now (in UploadFileForm) but easy to reintroduce.."
    )
    raise
print(f" - Memory footprint after loading Django: {resource.getrusage(resource.RUSAGE_SELF)[2] / 1024.0:.3f} MB")

from django.contrib.auth.models import User
from django.core import management
from django.db import connection, transaction

from troggle.core.utils import get_process_memory
from troggle.parsers.imports import (
    import_caves,
    import_drawingsfiles,
    import_ents,
    import_loadpos,
    import_logbook,
    import_logbooks,
    import_people,
    import_QMs,
    import_survex,
    import_surveyscans,
)

if os.geteuid() == 0:
    # This protects the server from having the wrong file permissions written on logs and caches
    print("This script should be run as expo not root - quitting")
    exit()

expouser = settings.EXPOUSER
expouserpass = settings.EXPOUSERPASS
expouseremail = settings.EXPOUSER_EMAIL

expoadminuser = settings.EXPOADMINUSER
expoadminuserpass = settings.EXPOADMINUSERPASS
expoadminuseremail = settings.EXPOADMINUSER_EMAIL


def reinit_db():
    """Rebuild database from scratch. Deletes the file first if sqlite is used,
    otherwise it drops the database and creates it.
    Note - initial loading of troggle.sqlite will already have populated the models
    in memory (django python models, not the database), so there is already a full load
    of stuff known. Deleting the db file does not clear memory.
    """
    print("Reinitialising db ", end="")
    print(django.db.connections.databases["default"]["NAME"])
    currentdbname = settings.DATABASES["default"]["NAME"]
    if currentdbname == ":memory:":
        # closing connections should wipe the in-memory database
        django.db.close_old_connections()
        for conn in django.db.connections.all():
            print(" ! Closing another connection to db...")
            conn.close()
    elif django.db.connections.databases["default"]["ENGINE"] == "django.db.backends.sqlite3":
        if os.path.isfile(currentdbname):
            try:
                print(" - deleting " + currentdbname)
                os.remove(currentdbname)
            except OSError:
                print(
                    " ! OSError on removing: "
                    + currentdbname
                    + "\n ! Is the file open in another app? Is the server running?\n"
                )
                raise
        else:
            print(" - No database file found: " + currentdbname + " ..continuing, will create it.\n")
    else:
        print(f" - Attempting to nuke : {currentdbname}\n")
        # this is now completely failing to nuke MariaDB adequately, and it crashes when creating Area objects with a no null parent message
        # when null parents are explciitly allowed in the model.
        cursor = django.db.connection.cursor()
        try:
            cursor.execute(f"DROP DATABASE {currentdbname}")
        except:
            pass
        cursor.execute(f"CREATE DATABASE {currentdbname}")
        cursor.execute(f"ALTER DATABASE {currentdbname} CHARACTER SET=utf8")
        cursor.execute(f"USE {currentdbname}")
        print(f" - Nuked : {currentdbname}\n")

    print(" - Migrating: " + django.db.connections.databases["default"]["NAME"])

    if django.db.connections.databases["default"]["ENGINE"] == "django.db.backends.sqlite3":
        # with transaction.atomic():
        management.call_command("makemigrations", "core", interactive=False)
        management.call_command("migrate", interactive=False)
        management.call_command("migrate", "core", interactive=False)
    else:
        management.call_command("makemigrations", "core", interactive=False)
        management.call_command("migrate", interactive=False)
        management.call_command("migrate", "core", interactive=False)

    print(" - done migration on: " + settings.DATABASES["default"]["NAME"])
    print("users in db already: ", len(User.objects.all()))
    with transaction.atomic():
        try:
            print(" - Setting up expo user on: " + django.db.connections.databases["default"]["NAME"])
            print(f" - user: {expouser} ({expouserpass:.5}...) <{expouseremail}> ")
            user = User.objects.create_user(expouser, expouseremail, expouserpass)
            user.is_staff = False
            user.is_superuser = False
            user.save()
        except:
            print(" ! INTEGRITY ERROR user on: " + settings.DATABASES["default"]["NAME"])
            print(django.db.connections.databases["default"]["NAME"])
            print(" ! You probably have not got a clean db when you thought you had.\n")
            print(" ! Also you are probably NOT running an in-memory db now.\n")
            print("users  in db: ", len(User.objects.all()))
            print("tables in db: ", len(connection.introspection.table_names()))
            memdumpsql(fn="integrityfail.sql")
            django.db.connections.databases["default"]["NAME"] = ":memory:"
            # raise

    with transaction.atomic():
        try:
            print(" - Setting up expoadmin user on: " + django.db.connections.databases["default"]["NAME"])
            print(f" - user: {expoadminuser} ({expoadminuserpass:.5}...) <{expoadminuseremail}> ")
            user = User.objects.create_user(expoadminuser, expoadminuseremail, expoadminuserpass)
            user.is_staff = True
            user.is_superuser = True
            user.save()
        except:
            print(" ! INTEGRITY ERROR user on: " + settings.DATABASES["default"]["NAME"])
            print(django.db.connections.databases["default"]["NAME"])
            print(" ! You probably have not got a clean db when you thought you had.\n")
            print(" ! Also you are probably NOT running an in-memory db now.\n")
            print("users  in db: ", len(User.objects.all()))
            print("tables in db: ", len(connection.introspection.table_names()))
            memdumpsql(fn="integrityfail.sql")
            django.db.connections.databases["default"]["NAME"] = ":memory:"
            # raise


def memdumpsql(fn):
    """Unused option to dump SQL. Aborted attempt to create a cache for loading data"""
    djconn = django.db.connection
    from dump import _iterdump

    with open(fn, "w") as f:
        for line in _iterdump(djconn):
            f.write(f"{line.encode('utf8')}\n")
    return True


# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -


class JobQueue:
    """A list of import operations to run. Always reports profile times
    of the import operations in the same order.
    """

    def __init__(self, run):
        """Initialises the job queue object with a fixed order for reporting
        options during a run. Imports the timings from previous runs.
        """
        self.runlabel = run
        self.queue = []  # tuples of (jobname, jobfunction)
        self.results = {}
        self.results_order = [
            "date",
            "runlabel",
            "reinit",
            "caves",
            "people",
            "logbooks",
            "QMs",
            "scans",
            "survex",
            "drawings",
            "test",
        ]
        for k in self.results_order:
            self.results[k] = []
        self.tfile = "import_profile.json"
        self.htmlfile = "profile.html"  # for HTML results table. Not yet done.

    def enq(self, label, func):
        """Enqueue: Adding elements to queue"""
        self.queue.append((label, func))
        return True

    def loadprofiles(self):
        """Load timings for previous imports for each data import type"""
        if os.path.isfile(self.tfile):
            try:
                f = open(self.tfile, "r")
                data = json.load(f)
                for j in data:
                    self.results[j] = data[j]
            except:
                print(f"FAILURE parsing JSON file {self.tfile}")
                # Python bug: https://github.com/ShinNoNoir/twitterwebsearch/issues/12
            f.close()
        for j in self.results_order:
            self.results[j].append(None)  # append a placeholder
        return True

    def dellastprofile(self):
        """trim one set of data from the results"""
        for j in self.results_order:
            self.results[j].pop()  # delete last item
        return True

    def delfirstprofile(self):
        """trim one set of data from the results"""
        for j in self.results_order:
            self.results[j].pop(0)  # delete zeroth item
        return True

    def saveprofiles(self):
        """Save timings for the set of imports just completed"""
        with open(self.tfile, "w") as f:
            json.dump(self.results, f)
        return True

    def runqonce(self):
        """Run all the jobs in the queue provided - once"""
        print("** Running job ", self.runlabel, end=" to ")
        print(django.db.connections.databases["default"]["NAME"])
        jobstart = time.time()
        print(f"-- Initial memory in use {get_process_memory():.3f} MB")
        self.results["date"].pop()
        self.results["date"].append(jobstart)
        self.results["runlabel"].pop()
        self.results["runlabel"].append(self.runlabel)

        for runfunction in self.queue:
            start = time.time()
            memstart = get_process_memory()
            jobname, jobparser = runfunction
            # --------------------
            jobparser()  #  invokes function passed in the second item in the tuple
            # --------------------
            memend = get_process_memory()
            duration = time.time() - start
            # print(" - MEMORY start:{:.3f} MB end:{:.3f} MB change={:.3f} MB".format(memstart,memend, ))
            print(
                '\n*- Ended "',
                jobname,
                f'"  {duration:.1f} seconds + {memend - memstart:.3f} MB ({memend:.3f} MB)',
            )
            self.results[jobname].pop()  # the null item
            self.results[jobname].append(duration)

        jobend = time.time()
        jobduration = jobend - jobstart
        print(f"** Ended job   {self.runlabel}  -  {jobduration:.1f} seconds total.")
        return True

    def append_placeholders(self):
        """Ads a dummy timing for each option, to fix off by one error"""
        for j in self.results_order:
            self.results[j].append(None)  # append a placeholder

    def run_now_django_tests(self, n):
        """Runs the standard django test harness system which is in troggle/core/TESTS/tests.py"""
        management.call_command("test", verbosity=n)
        django.db.close_old_connections()

    def run(self):
        """Initialises profile timings record, initiates relational database, runs the job queue saving the imported data as an SQL image and saves the timing profile data."""
        self.loadprofiles()
        print(
            "-- start   ",
            django.db.connections.databases["default"]["ENGINE"],
            django.db.connections.databases["default"]["NAME"],
        )
        self.runqonce()
        if settings.DATABASES["default"]["NAME"] == ":memory:":
            memdumpsql("memdump.sql")  # saved contents of in-memory db, could be imported later..
        self.saveprofiles()
        return True

    def showprofile(self):
        """Prints out the time it took to run the jobqueue"""
        for k in self.results_order:
            if k == "test":
                break
            elif k == "date":
                print("     days ago ", end=" ")
            else:
                print("%10s (s)" % k, end=" ")
            percen = 0
            r = self.results[k]

            for i in range(len(r)):
                if k == "runlabel":
                    if r[i]:
                        rp = r[i]
                    else:
                        rp = "      - "
                    print("%8s" % rp, end=" ")
                elif k == "date":
                    # Calculate dates as days before present
                    if r[i]:
                        if i == len(r) - 1:
                            print("    this", end=" ")
                        else:
                            # prints one place to the left of where you expect
                            if r[len(r) - 1]:
                                s = r[i] - r[len(r) - 1]
                            elif r[len(r) - 2]:
                                s = r[i] - r[len(r) - 2]
                            else:
                                s = 0
                            days = (s) / (24 * 60 * 60)
                            print(f"{days:8.2f}", end=" ")
                elif r[i]:
                    print(f"{r[i]:8.1f}", end=" ")
                    if i == len(r) - 1 and r[i - 1]:
                        percen = 100 * (r[i] - r[i - 1]) / r[i - 1]
                        if abs(percen) > 0.1:
                            print(f"{percen:8.1f}%", end=" ")
                else:
                    print("      - ", end=" ")
            print("")
        print("\n")
        return True


def usage():
    """Prints command line options, can print history of previous runs with timings"""
    print(
        """Usage is 'python databaseReset.py <command> [runlabel]'
             where command is:
             test      - testing... imports people and prints profile. Deletes nothing.
             profile   - print the profile from previous runs. Import nothing. 
                        -  del       - deletes last entry
                        -  delfirst  - deletes first entry

             reset     - normal usage: clear database and reread everything from files
             
             init      - initialisation. Automatic if you run reset.
             caves     - read in the caves (must run first after initialisation)
             people    - read in the people from folk.csv (must run after 'caves')
             logbooks  - read in the logbooks
             QMs       - read in the QM csv files (older caves only)
             scans     - the survey scans in all the wallets (must run before survex)
             drawings  - read in the Tunnel & Therion files - which scans the survey scans too
             survex    - read in the survex files - all the survex blocks and entrances x/y/z 
             ents       - read just the entrances x/y/z (must run after survex)

             dumplogbooks - Not used. write out autologbooks (not working? use http://localhost:8000/controlpanel )
             logbook   - read a single logbook. Defautl set in python code

             and [runlabel] is an optional string identifying this run of the script
             in the stored profiling data 'import-profile.json'
             
             caves and logbooks must be run on an empty db before the others as they
             set up db tables used by the others.

             Note that running the subfunctions will not produce a consistent website 
             - only the full 'reset' does that.
             """
    )


if __name__ == "__main__":

    if os.geteuid() == 0:
        print("Do not run as root or using sudo - file permissions for cache files and logs will break")
        print("Aborting run.")
        exit()

    if sys.getfilesystemencoding() != "utf-8":
        print("UTF-8 is NOT the default file encoding. You must fix this.")
        print(f"- {sys.getdefaultencoding()=}")
        print(f"- {sys.getfilesystemencoding()=}")
        print(f"- {locale.getdefaultlocale()=}")
        print(f"- {locale.getpreferredencoding()=}")
        print("Aborting run.")
        exit()

    if len(sys.argv) > 2:
        runlabel = sys.argv[len(sys.argv) - 1]
    else:
        runlabel = None

    jq = JobQueue(runlabel)

    if len(sys.argv) == 1:
        usage()
        exit()
    elif "init" in sys.argv:
        jq.enq("reinit", reinit_db)
    elif "ents" in sys.argv:
        jq.enq("survex", import_ents)
    elif "test2" in sys.argv:
        jq.enq("QMs", import_QMs)
        jq.enq("drawings", import_drawingsfiles)
        jq.enq("survex", import_survex)
    elif "caves" in sys.argv:
        jq.enq("caves", import_caves)
    elif "logbooks" in sys.argv:
        jq.enq("logbooks", import_logbooks)
    elif "logbook" in sys.argv:
        jq.enq("logbooks", import_logbook)  # default year set in imports.py
    elif "people" in sys.argv:
        jq.enq("people", import_people)
    elif "QMs" in sys.argv:
        jq.enq("QMs", import_QMs)
    elif "reset" in sys.argv:
        jq.enq("reinit", reinit_db)
        jq.enq("caves", import_caves)
        jq.enq("people", import_people)
        jq.enq("scans", import_surveyscans)
        jq.enq("logbooks", import_logbooks)
        jq.enq("QMs", import_QMs)
        jq.enq("drawings", import_drawingsfiles)
        jq.enq("survex", import_survex)
    elif "scans" in sys.argv:
        jq.enq("scans", import_surveyscans)
    elif "survex" in sys.argv:
        jq.enq("survex", import_survex)
    elif "loadpos" in sys.argv:
        jq.enq("survex", import_loadpos)
    elif "drawings" in sys.argv:
        jq.enq("drawings", import_drawingsfiles)
    elif "profile" in sys.argv:
        if runlabel == "del":
            jq.loadprofiles()
            jq.dellastprofile()
            jq.dellastprofile()  # twice because loadprofiles adds a dummy
            jq.showprofile()
            jq.saveprofiles()
        if runlabel == "delfirst":
            jq.loadprofiles()
            jq.dellastprofile()  # remove the dummy
            jq.delfirstprofile()
            jq.showprofile()
            jq.saveprofiles()
        else:
            jq.loadprofiles()
            jq.showprofile()
        exit()
    elif "help" in sys.argv:
        usage()
        exit()
    else:
        usage()
        print(f"{sys.argv[1]} not recognised as a command.")
        exit()

    jq.run()
    jq.showprofile()